hidapi VS OpenHMD

On the computer, write a program (the "host" or "server") that talks over the raw HID protocol. It is open ended how you do this, but IME both the Python pyhidapi and C hidapi were simple enough to get something started. The trickiest part might be getting the program to run with the permissions to talk the raw HID protocol (at least on Linux, there was some udev rule to set up). Read the documentation for these libraries for details about how to get set up.

now I'm trying to figure out what to do with the next step for building my own host. Am I supposed to put this test file in my qmk code? or run it on my computer as it's own separate program? https://github.com/libusb/hidapi/blob/master/hidtest/test.c

HIDAPI

I did star https://github.com/libusb/hidapi too, as well as this GitHub.

I used ffigen to get the bindings from hidapi.h and cleaned up quite some of the resulting output to make it Dart compatible. Quite some typedef and some minor stuff.

I figured out the problem with JoyconLib! The hidapi.dll file was broken so I replaced it with a new file from https://github.com/libusb/hidapi/releases and not it's working.

https://github.com/libusb/hidapi#prerequisites

Install the HID API, instructions here: https://github.com/libusb/hidapi

Inspired by https://askubuntu.com/questions/699138/logitech-k380-bluetooth-keyboard-make-function-keys-default and https://github.com/libusb/hidapi, it's almost re-implementation of hidapi actually, with dart:ffi calling the syscall directly.

I modified the AUR package for openhmd to build your repo/branch instead. git+https://github.com/thaytan/OpenHMD.git#branch=rift-kalman-filter But the tracking is just as jittery as before and it's still just 3DOF. Is there any special compiler flags or something that need to be activated for 6DOF to work, or is it more likely that I just installed it wrong somehow?

Yes. Was developed by the same guy, in fact: https://github.com/thaytan/OpenHMD/tree/rift-kalman-filter

if they wanna stay with Oculus, they could give the CV1 a try, assuming they know how to setup the drivers for it to work on Linux. Thaytan on github has been working on and improving 6DOF tracking for the CV1 for a while now but is needing help with the tracking part of it to fix a few kinks. You can easily okay Beat Saber with it no problem. Should give it a look if you are interested. Thaytan has some youtube videos as well of some gameplay with it. The branch on github you would want to use is the rift-correspondence-search one. https://github.com/thaytan/OpenHMD/tree/rift-correspondence-search

Oculus Rift DK2/CV1: Try my positional tracking branches of OpenHMD (https://github.com/thaytan/OpenHMD/tree/rift-kalman-filter and https://noraisin.net/diary/?page_id=1048)

Slightly out of date tutorial: https://noraisin.net/diary/?page_id=1048 Github repo for the experimental support: https://github.com/thaytan/OpenHMD

I re-measured distortion parameters on my CV1 last month (https://github.com/thaytan/OpenHMD/commit/46d164a9) and they made things much clearer for me... maybe those parameters aren't right for your headset? Unfortunately it's a really hard thing to get a photo of for me to tell what you're seeing on your end.

OpenHMD Rift Kalman Filter: https://github.com/thaytan/OpenHMD/tree/rift-kalman-filter

https://github.com/ChristophHaag/SteamVR-OpenHMD with https://github.com/thaytan/OpenHMD/tree/rift-kalman-filter would work for now getting the CV1 working positionally tracked on Linux (currently in branch, still experimental but usable!) There are step by step install instructions on https://noraisin.net/diary/?p=1033